New Models

Added support for the 16k models

Rate Limit Pause Period

Update to one second

Tweak Rate Limiting

Small bug. Adjusted a bit, still need to clean up a bit more.

Avoid Rate Limiting

Using an actor to avoid getting rate-limited in a thread-safe manner

Adjusted Token Heuristic

We now skip obviously doomed files and leave a warning about files unlikely to be returned back in their entirety by DoccGPT

Adopt SwiftLog

Replaced all print statements

Use Chat API

Works just as well and gives us access to better, cheaper models

Options and tweaks

Add warning for long files and the ability to skip them. Working on more robust workarounds.

Swift Concurrency

Improves performance by parallelizing the write operations using Swift Concurrency

Initial release

0.1.0

Update README.md